For sponsored programmes and research administration
Effort certification is a personal attestation by the person who did the work. Allowability is a determination your office holds under federal rule. Scientific content belongs to the investigator. What can be carried is the assembly, the chasing and the reconciliation that consume a sponsored programmes office in the seventy-two hours before every deadline.
The deadline belongs to the sponsor and is absolute. The proposal arrives from a principal investigator who has been teaching, running a lab and reviewing for a journal, and who is not late out of disregard — the grant is the fourth job.
So the office receives a near-complete proposal a day before submission, with a budget that needs building, a justification that needs formatting, biosketches in the wrong template, subrecipient commitment letters that have not arrived, and compliance certifications nobody has signed. Then it does the same thing again for six other proposals that week.
The office has responsibility and no authority. A principal investigator does not report to research administration, has tenure, brings in money, and can escalate to a dean. Every deadline conversation is a negotiation between people with asymmetric leverage.
Post-award is quieter and larger. Subaward monitoring, cost transfers, effort certification cycles, no-cost extensions, carryforward requests, closeout — all running for years across a portfolio, all with the office accountable and the investigator holding the facts.
And the audit arrives years later, asking about a cost transfer nobody remembers, on an award whose personnel have all moved on, with the institution answerable regardless.
A proposal becomes the office’s responsibility only when the investigator hands it over, which is systematically too late to do the assembly carefully — so completeness is achieved by heroics or not at all.
That is not fixable by asking investigators to submit earlier. It has been asked for thirty years and the incentives have not changed: the deadline that matters to a principal investigator is the sponsor’s, and internal deadlines are advisory in practice.
So the useful move is to make the assembly fast rather than to make the handoff earlier. Every requirement for that sponsor and that mechanism, checked at the moment the material arrives. Every missing item named at once. Templates and formats applied from the institution’s own library. The subrecipient commitment chased from the day the collaboration was named rather than from the day before submission.
Post-award the band is monitoring and reconciliation: which subawards have reports outstanding and how old, which effort certifications are open, which cost transfers lack contemporaneous documentation, which awards are approaching an end date without a closeout started.
What never moves: effort certification, which is a personal attestation and cannot be delegated to anybody; allowability determination, which your office holds under federal rule; any decision about a proposal’s scientific content or its budget adequacy; and the institutional signature. Those are the job.
Proposals complete at internal deadline rather than at submission — measured by share of proposals with every required element present at the internal deadline, before and after.
Subrecipient commitments outstanding at submission — measured by count of proposals submitted with a commitment document still missing.
Hours in the final seventy-two before a deadline — measured by office hours logged in the three days before submission, per proposal, sampled the same way.
Effort certifications open past their cycle — measured by count and age of open certifications, by department.
Cost transfers lacking contemporaneous documentation — measured by count identified within thirty days versus at audit.
Subaward reports outstanding — measured by count and age by subrecipient, which most offices track partially.
Awards approaching end date with no closeout started — measured by count inside a stated window of the end date.
any effort certification, which is a personal attestation and is not delegable to anybody; any allowability, allocability or reasonableness determination; any decision about a proposal’s scientific content, budget adequacy or competitiveness; and the institutional signature. Nothing here signs, certifies or determines anything on behalf of the institution or an individual.
It reads and writes the electronic research administration system, the financial system and the effort system through documented interfaces, and creates no second record of an award. A second record of an award is precisely what an audit will find and ask about.
Requirement lists come from the sponsor’s published guidance as your office already interprets it, plus your institution’s own policy. Nothing interprets a sponsor’s guidance — where two of your own sources disagree, that is reported to your officer rather than resolved.
Investigator-facing surfaces have to work for somebody checking a message between experiments on a phone, in two minutes, and any surface that assumes a desk session will not be used.
Where a sponsor system has no documented interface, that is a limitation and the step stays manual. Automating a submission portal is refused: a failed submission against an absolute deadline is unrecoverable.
Effort certification is a personal attestation by the individual who performed the work. It cannot be delegated to a department administrator, to a system, or to us — and a vendor offering to make effort certification easier by doing any part of the attestation is offering a federal problem.
The audit file is what this produces. Every action carries what was done, under which grant of authority, by which actor, at what time, against which award, with what result — created during the work rather than reconstructed for an auditor years later, when the personnel have gone.
Institutional and investigator data stays inside your tenancy, on your retention schedule, exportable by you, and is not used to train anything serving another organisation. Unfunded proposal content in particular is competitively sensitive and never becomes training material.
And on assurance: an independent SOC 2 Type II attestation is in progress and no report exists yet.
The authorised organisational representative signs, and nothing here changes that. What changes is what is in front of them when they do.
Internal audit should review the record format against a real prior audit finding. If the record would not have answered that finding, it is not the right record and that is worth knowing now.
Research compliance will want the effort boundary in writing, because it is the one an efficiency argument erodes first. It is enumerated: the cycle is chased and the attestation is untouched, at any tier.
And counsel should confirm the handling of unfunded proposal content, which is competitively sensitive and belongs to investigators as much as to the institution.
One sponsor and mechanism’s proposals across a single deadline cycle, measured for completeness at internal deadline and for hours in the final seventy-two, with no authority to submit or sign anything.
Narrow to one mechanism because requirement lists are mechanism-specific, and a general measurement across sponsors produces an average that describes nobody.
The output is two numbers most offices do not have: what share of proposals were actually complete at the internal deadline, and how many office hours the final three days consumed. Both are useful in a resourcing conversation with a dean whatever happens next.
If it continues, the first grant covers completeness checking and the subrecipient commitment chase for that one mechanism — with certification, allowability and signature outside the scope document.
Most will not, and a design that requires them to is already failed. The band that matters runs between the office and the department administrator, not through the investigator — the subrecipient chase, the completeness list, the format application. Where an investigator is contacted, it has to work in two minutes on a phone between other obligations, and if the measurement shows they are ignoring it, the honest response is to route around them rather than to ask them again.
Because the attestation is personal and non-delegable — a vendor touching any part of it is a federal problem regardless of how carefully it is done. What can move is everything around it: which certifications are open, how old, in which department, and whether the underlying cost transfers were documented contemporaneously. That is where most findings actually originate, and it is chaseable without anybody attesting to anything on somebody else’s behalf.
Agreed, which is why it is refused outright rather than carefully engineered. A failed submission against an absolute deadline is unrecoverable and the blame is irrelevant to the investigator who lost a year. Submission stays entirely with your office. What is carried is everything that determines whether the material is ready when your office goes to submit.
It is, and unfunded content is the most sensitive of all — an idea that did not get funded this cycle is being resubmitted next cycle. It stays in your systems, on your retention schedule, exportable by you, and is never training material at any tier. If your counsel wants that as a contractual term rather than an architectural claim, it should be both.